OxML Lecture: Geometric Deep Learning Explained
mmbronstein · x · 2026-07-16
Sharing a lecture from OxML 2026 focused on Geometric Deep Learning.
The post notes that Petar Veličković introduced "Grids, Groups, Graphs, Geodesics and Gauges," reviewing the core ideas from the Geometric Deep Learning book and framework: most real-world learning problems aren't arbitrary function fitting, but involve structures and symmetries derived from the physical world.
The summary also emphasizes the framework's connection to Felix Klein's Erlangen Programme, suggesting that different learning architectures can be viewed as distinct ways of handling geometric symmetries.
More from Research
- OpenAI says long-horizon models need safety and alignment checks across full action sequences — rhiever · 2026-07-22
- A Reddit user proposes a consistency LoRA to keep anime and game scenes visually stable — ThirdWorldBoy21 · 2026-07-22
- Graph workload 854.graph500 enters SPEC CPU 2026 as a new CPU benchmark — Prof_DavidBader · 2026-07-22
- BlackboxNLP 2026 is recruiting extra reviewers after a high submission volume — hanjie_chen · 2026-07-22
- AWS shows self-distilled reasoning can preserve math and coding skills during SFT — AWS ML Blog · 2026-07-22
- UI2App shows screenshot fidelity still lags real interaction recovery — Grace Man Chen · 2026-07-22